ComfyUI plug-in summary

Recommended commonly used plug-ins

1. AIGODLIKE-COMFYUI-TRANSLATION (Translate all the processes in the editor into Chinese,)

ComfyUI's translation plug-in is a must-installed plug-in. It can translate all English in the node editor into Chinese, and of course it also supports languages ​​in other countries. The official introduction currently supports Chinese simplified Chinese, Chinese translation, Japanese, Korean, etc. After all, sometimes, all English is sometimes easily confused and cannot distinguish nodes and related configuration parameters, so it is still necessary to convert Mandarin (you can ignore good English)

GitHub address: https://github.com/AIGODLIKE/AIGODLIKE-COMFYUI-TRANSLATION

2.comfyui-workspace-manager

ComfyUI Workflow Manager, this plug-in is used to manage ComfyUI workflows. Under normal circumstances, there is no workflow switching. If you want to open another workflow (check the previous configuration or run it at the same time), you need to close the current find workflow file and reload it. This plug-in can edit the notes workflow, and can open multiple workflows and switch back and forth. It can also be understood as the browser's tab bar.

GitHub address: https://github.com/11cafe/comfyui-workspace-manager
3. SDXL Prompt Styler

The function of SDXL stylized prompt words is equivalent to presetting various styles of prompt words in advance through the plug-in. We only need to select the specified style option and will automatically be added to the prompt words we finally output to the sampler. You don’t need to check certain styles or what styles are there every time, it’s more humane and convenient. No need to match a special Lora model to generate the style pictures we want. This plug-in has preset style options of more than 100+, which actually has 3 built-in json configuration files. We can also modify the json file ourselves to add styles that are not available. Although the name is SDXL stylized prompt word, it is also compatible with the SD1.5 model.

GitHub address: https://github.com/twri/sdxl_prompt_styler

4. ComfyUI-WD14-Tagger

Reverse push plug-in, this plug-in is also very popular on SD. It is used to reverse the prompt words of existing pictures, helping us better optimize our prompt words and learn from the results of reverse push, but it cannot be copied all, because in some scenarios, it may not be completely reversed correctly. There are 4 main parameters, and the following are explained to you one by one:

Model: Switch the anti-heap prompt word model (recommended covnet_tagger_v2 model)

threshold: refers to the label being regarded as a valid score. The lower the value, the more detailed the prompt word generated.

character_threshold: refers to the score that the character label is considered valid (usually no adjustment is made)

exclude_tag: Exclude the prompt words written (cannot be filled in by default)

If it is not used in the process in a node, you can also right-click WD14-Tagger to directly reverse the heap for the generated image, and the final prompt word will be returned in the form of a pop-up window.

GitHub address: https://github.com/pythongosssss/ComfyUI-WD14-Tagger

5. ComfyUI-Crystools

A ComfyUI tool set plugin, with this set, you can view CPU/GPU/memory resource consumption, image production progress and used time, metadata and compare two images, compare two JSONs, display any value to console/display, pipelines, etc.! This provides better nodes to load/save images, previews, etc. and to view the "hidden" data without loading a new workflow. What I prefer is that you can see the computer resource consumption and the progress of the picture in real time.

GitHub address: https://github.com/crystian/ComfyUI-Crystools

6. ComfyUI_UltimateSDUpcale

Literally, a plug-in for high-definition amplification of generated images. Friends who often use SD should have a good experience. Generally, existing images are processed again, because sometimes the images may have blurry or unclear problems, so you can use this plug-in to repair and optimize them.

GitHub address: https://github.com/situ/ComfyUI_UltimateSDUpcale

7. OneButtonPrompt

Chinese translation: One-click (generated) prompt word. Yes, with it you no longer need to type the prompt word by yourself. It is especially suitable for beginners or friends who don’t know how to write the prompt word. It can automatically write propt for you, and the generated pictures may be better than you write them yourself. (Lazy Artifact)

Steps to use:

Generate full prompt by clicking a button

Support TXT2IMG, IMG2IMG, ControlNET, inpainting and latent couple

Supports selection of topic types, artists, and image types

GitHub address: https://github.com/AIrjen/OneButtonPrompt

8. SuperBeasts
This plugin enhances the dynamic range and visual appeal of the input image.

Node specific functions:

Adjust the intensity of shadows, highlights, and overall HDR effects.

Apply gamma correction to control overall brightness and contrast.

Enhance contrast and color saturation for a more vivid effect.

Maintain color accuracy by processing images in the LAB color space, and targeted adjustments are made using brightness-based masks.

Mix the adjusted brightness with the original brightness for a balanced effect.

The functions and effects of parameters:

HDR intensity (default: 0.5, range: 0.0 to 5.0, adjustment step size: 0.01):

Controls the overall intensity of high dynamic range (HDR) effects in pictures. Higher values ​​will make the HDR effect more prominent and make the contrast between light and dark in the image more vivid.

Shadow intensity (default: 0.25, range: 0.0 to 1.0, adjustment step size: 0.01):

Used to adjust the depth of the shadow part in the image. The higher the value, the deeper the shadow, which enhances the contrast of the image and makes the details more distinct.

Highlight intensity (default: 0.75, range: 0.0 to 1.0, adjustment step size: 0.01):

Adjusts the brightness of the bright parts in the image. By increasing this value, the highlights will be brighter, further increasing the contrast and making the image more vivid.

Gamma intensity (default: 0.25, range: 0.0 to 1.0, adjustment step size: 0.01):

Determines the degree of gamma correction of the image. Gamma correction affects the overall brightness and contrast of the image. The higher the value, the brighter the overall image and the stronger the contrast.

Contrast (default: 0.1, range: 0.0 to 1.0, adjustment step size: 0.01):

Used to enhance the overall contrast of the image. By increasing this value, the contrast of the image will be more significant, making the details of the image clearer and more prominent.

Color enhancement (default: 0.25, range: 0.0 to 1.0, adjustment step size: 0.01):

Improves the color saturation of the image. As the value increases, the color of the image becomes more vivid and vivid.

GitHub address: https://github.com/SuperBeastsAI/ComfyUI-SuperBeasts

9. ComfyUI WebGUI: GitHub – ltdrdata/ComfyUI-Manager (This plugin is currently equipped with desktop installation, but there is no Chinese version)

Workflow configuration, plug-in installation, model management, etc. through WebUI.

GitHub address: https://github.com/ltdrdata/ComfyUI-Manager

10. ComfyUI-Inspyrenet-Rembg

Remove the background of the picture

GitHub address: https://github.com/john-mnz/ComfyUI-Inspyrenet-Rembg
11. ComfyUI-InstantID

Convert a character photo into a cartoon photo

GitHub address: https://github.com/ZHO-ZHO-ZHO/ComfyUI-InstantID
12. ComfyUI_CatVTON_Wrapper

For character dressing

GitHub address: https://github.com/chflame163/ComfyUI_CatVTON_Wrapper

13. ComfyUI_LayerStyle

Implementing Photoshop-like synthetic layer and mask functions

GitHub address: https://github.com/chflame163/ComfyUI_LayerStyle

14. ComfyUI-Inpaint-Nodes

Remove the items specified in the picture and perform natural repairs based on the current surrounding environment of the picture.

GitHub address: https://github.com/Acly/comfyui-inpaint-nodes

15. rgthree-comfy

Better orchestrate and manage workflows.

GitHub address: https://github.com/rgthree/rgthree-comfy

16. ComfyUI_IPAdapter_plus

Used to fuse two images. For example: Change one person's face to another person's body.

GitHub address: https://github.com/cubiq/ComfyUI_IPAdapter_plus

17. ComfyUI_Controlnet_AUX

Anime style, cyberpunk style.

GitHub address: https://github.com/Fannovel16/comfyui_controlnet_aux

Plugin Summary

1. Workspace

2. Comprehensive

3. Control

4. Video

5. Others